wattlebird / ditech

Ditech competition http://research.xiaojukeji.com/competition/main.action?competitionId=DiTech2016

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Ditech Competition 2016

This repository is for the code of my competing Ditech Competition 2016. The final rank is 269/775. The top score is 0.290618 (first place score is 0.245202).

How to use?

Download data in the same folder, remove "_test" in all the test data file filename.

How to play?

Jupyter notebook "Feature_extraction_playground.ipynb" is the scratch paper of feature extraction and data checking. "Model_training_playground.ipynb" is the scratch paper for model selection and parameter tunning.

Run python feature_generation.py first then python main.py to generate result.csv.

About

Ditech competition http://research.xiaojukeji.com/competition/main.action?competitionId=DiTech2016


Languages

Language:Jupyter Notebook 97.4%Language:Python 2.6%